Peri-Implantitis: How Common It Really Is, What Causes It, and What the Evidence Says Prevents It

Peri-implantitis is the reason a healthy implant can start losing bone years after it integrated, and the reason I spend part of every consultation on hygiene and recall rather than on the surgery. Before any prevalence figure makes sense you need the definitions, and the field only agreed on them in 2017. The World Workshop on the Classification of Periodontal and Peri-Implant Diseases, reported by Berglundh and 24 co-authors in 2018, set out three states. Peri-implant health is the absence of redness, bleeding on probing, swelling and pus; it can exist around an implant that has already lost some bone, and no single probing depth defines it. Peri-implant mucositis is inflammation of the soft tissue — its cardinal sign is bleeding on gentle probing — without progressive bone loss. Peri-implantitis is inflammation plus progressive loss of the supporting bone.

Renvert's accompanying paper supplied the case definitions used in the clinic. Health: no signs of inflammation and no bone loss beyond initial healing. Mucositis: bleeding or a drop of blood within 30 seconds of probing, redness or swelling, and no bone loss beyond initial healing. Peri-implantitis: signs of inflammation, radiographic bone loss after initial healing, and probing depths that have increased compared with the measurement taken when the crown was fitted. When no earlier radiograph exists, a bone level of 3 mm or more from the implant shoulder together with bleeding and probing depths of 6 mm or more is taken as indicative of disease. Note what that last clause implies: without a baseline, the diagnosis is a guess. The Workshop's recommendation that every clinician record a radiograph and probing depths at the moment the prosthesis is delivered is the single cheapest thing in this article, and I have made it routine.

Schwarz, Derks, Monje and Wang's review for the same Workshop describes how the disease behaves once it starts. Onset can be early in follow-up, and progression is non-linear and accelerating rather than steady. Lesions are typically larger than a comparable periodontitis lesion around a tooth, and surgical entry usually reveals a circumferential crater of bone loss around the implant. Progressive bone loss with no inflammation at all — the picture that would support a purely mechanical or foreign-body explanation — was, they concluded, rare.

The honest answer is that it depends on how you count, and the range is uncomfortably wide. Derks and Tomasi's 2015 systematic review of eleven studies found mucositis prevalence ranging from 19% to 65% and peri-implantitis from 1% to 47% depending on the definition used, with weighted means of 43% for mucositis and 22% for peri-implantitis. Their meta-regression found what you would expect: prevalence rose with the time implants had been in function and fell as the bone-loss threshold used to define disease rose. Atieh's 2013 review of nine studies, 1,497 patients and 6,283 implants, put peri-implantitis at 18.8% of patients and 9.6% of implants, and found a higher frequency — 36.3% — in smokers.

Lee's 2017 meta-analysis of 47 studies with at least three years of follow-up is the largest. Peri-implantitis affected 9.25% of implants and 19.83% of patients; mucositis affected 29.48% of implants and 46.83% of patients. The patient-level figure is the one to remember: about one implant patient in five develops peri-implantitis somewhere in the mouth within the follow-up windows studied, and nearly half develop mucositis. A 2022 meta-analysis by Diaz and colleagues covering 57 studies through 2021 landed in the same place, at 19.53% of patients and 12.53% of implants, and repeated the warning that the estimates remain highly variable even when restricted to a single case definition.

The most sobering data come from the Swedish population sample I cite in the survival-rate review. Derks and colleagues examined 588 randomly selected patients nine years after treatment. Using a threshold of any bone loss beyond 0.5 mm with bleeding or pus, 45% had peri-implantitis. Using a threshold of more than 2 mm — moderate or severe disease — 14.5% did. Higher odds of the moderate-to-severe form were found in patients with periodontitis, patients with four or more implants, implants of certain brands, prosthetic work delivered by general practitioners rather than referral clinicians, implants in the lower jaw, and crowns whose margin sat within 1.5 mm of the bone at baseline. Several of those are design decisions. That is the part of the list I can act on.

The cause is bacterial plaque, and that is not a hypothesis. Heitz-Mayfield and Salvi's review for the World Workshop describes the experimental model in which volunteers with implants stop cleaning for three weeks: an inflammatory infiltrate develops in the connective tissue beside the implant, and it develops in direct response to biofilm. Salvi's 2012 study of fifteen such volunteers found something more specific — the soft tissue around implants mounted a stronger inflammatory response to the same three weeks of plaque than the gum around neighbouring teeth did, with higher levels of the enzyme MMP-8 in the fluid around implants throughout. Both tissues recovered at the biomarker level once cleaning resumed, but three weeks of good hygiene did not return either to its starting clinical condition. Inflammation around an implant starts faster and leaves more slowly than it does around a tooth.

Whether mucositis becomes peri-implantitis is where the risk factors act, and Schwarz's Workshop review grades them. Strong evidence: a history of chronic periodontitis, poor plaque control, and no regular maintenance after the implant is placed. Inconclusive: smoking and diabetes — not because they are harmless, but because the studies are confounded; the diabetes review goes into why. Limited but real: residual cement under the gum after a cemented crown, a lack of firm keratinised tissue around the implant, and implant positions that make cleaning physically difficult. Hämmerle and Tarnow's companion review adds the anatomical background: thin soft tissue, malpositioned implants, mechanical overload and the natural resorption of the ridge after extraction all contribute to the tissue deficiencies in which disease takes hold.

The cement finding deserves its own paragraph because it is so avoidable. Wilson's 2009 study used a dental endoscope to look under the gum of 42 implants with signs of disease and 20 healthy controls in one periodontal practice. Excess cement was found at 34 of the 42 diseased implants — 81% — and at none of the controls. Thirty days after the cement was removed, 25 of 33 treated sites had no clinical or endoscopic sign of inflammation. A cemented implant crown can be done well, but a screw-retained one cannot leave cement behind, and that is the main reason I design screw-retained whenever the position allows. Smoking's effect on the implant itself is unambiguous even where its link to peri-implantitis specifically is debated: Chrcanovic's 2015 meta-analysis of 107 studies counted 6.35% failures among 19,836 implants in smokers against 3.18% among 60,464 in non-smokers, with more post-operative infections and more bone loss. Smoking and implants has its own page.

Maintenance is the intervention with the clearest evidence, and the effect size is large. Costa's 2012 five-year study followed 80 patients who all had mucositis at baseline. Among those who received preventive maintenance visits over the five years, 18.0% went on to develop peri-implantitis. Among those who did not, 43.9% did. Same disease, same starting point, and the difference was whether anyone was looking. Monje's 2016 meta-analysis of ten clinical trials confirmed the pattern: the interval between maintenance visits influenced the incidence of peri-implantitis, patients with a history of periodontitis were more affected, and the authors argued for a minimum recall interval of five to six months, tailored to risk. Their conclusion is the one I give patients in writing: implant therapy is not finished when the tooth goes in.

The 2015 European consensus on primary prevention, led by Jepsen with the same group of authors, turned the evidence into instructions. Bleeding on probing is the key measure that separates health from disease. Patient-administered mechanical cleaning — a manual or powered toothbrush used properly, with interdental cleaning — is an effective preventive measure, and professional cleaning with hygiene instruction reduces signs of inflammation. Adjuncts such as antiseptic rinses, local or systemic antibiotics and air-abrasive devices did not improve on professional mechanical cleaning for mucositis. Smoking was identified as the modifiable patient factor and excess cement as the local one. None of this is exotic. It is a hygienist, a probe, a radiograph and a calendar.

Design is the part prevention that happens before the patient ever holds a toothbrush. The Swedish finding that crown margins within 1.5 mm of bone carried higher odds of disease, and the Workshop's note that implant positions which obstruct cleaning are a risk indicator, both argue for planning the restoration before the implant rather than after it. A full-arch bridge whose underside cannot be reached with a brush or a water flosser is a peri-implantitis case waiting to happen, however good the surgery was. Our cleaning guide for full-arch teeth exists because of that, and the general care guide covers single implants.

Once bone has been lost, the literature is more modest, and I would rather you read that here than discover it later. Heitz-Mayfield and Mombelli's 2014 systematic review of 43 publications on peri-implantitis therapy found no trials comparing surgical with non-surgical treatment and too much heterogeneity to pool anything. Using a composite success criterion at twelve months, successful outcomes were reported in anywhere from 0% to 100% of patients across nine studies, with a majority of patients successful in seven of them. Recurrence, progression and implant loss despite treatment were all reported. What the successful protocols had in common was structure: a pre-treatment phase to control the patient's overall hygiene and risk factors, cause-related therapy at the implant, and a maintenance phase afterwards.

In practice that means non-surgical debridement and hygiene correction for early disease, and surgical access — opening the site, cleaning and decontaminating the implant surface, and either reshaping the bone or attempting to regenerate it — for established craters, with the honest caveat that regeneration around a contaminated titanium surface is less predictable than around a tooth. Mucositis, by contrast, is reversible, and treating it is the most effective peri-implantitis treatment there is, because it is the stage before bone has gone. Patients considering either route can read the consent forms we use for non-surgical and surgical treatment before an appointment. If disease has already cost an implant, this is what happens next.

Every recommendation above has a corresponding habit. A radiograph and six-point probing are recorded on the day the final prosthesis is delivered, so that a change five years later is measured against a number rather than a memory. Restorations are screw-retained wherever the implant position allows it, and where a crown must be cemented the margin is kept where excess can be seen and removed. Full-arch bridges are designed with cleanable undersides and the patient is shown, with the bridge in hand, how to reach them. Recall is set at three to six months according to the patient's periodontal history, smoking status and the number of implants, and it is written into the treatment plan rather than suggested at the door.

Patients with a history of periodontitis get a specific conversation, because that is the strongest risk factor in the literature and the one most often glossed over when the missing teeth were lost to gum disease in the first place. Implants are not immune to the bacteria that cost you your teeth; if anything, Salvi's data suggest the tissue around them reacts more strongly. The plan for such patients includes periodontal control before implants are placed, closer recall after, and a frank discussion about smoking. Bruxism is checked too, because heavy loading contributes to the tissue deficiencies Hämmerle describes; Chrcanovic's 2016 series found implant failure rates of 13.0% in patients with bruxism against 4.6% in those without, and grinding has its own page.

Three developments would change practice if they mature. The first is diagnostic: a probe and a radiograph record destruction that has already happened, and Alassy's 2019 review of biomarkers in peri-implant crevicular fluid — interleukin-1β, tumor necrosis factor-α and MMP-8 among them — suggests that a chairside test could one day report disease activity before bone is lost, though the review is clear that prediction of progression is not yet demonstrated. The second is a settled case definition applied uniformly, so that the next decade of prevalence studies can be pooled without the tenfold spread that the current ones show. The third is the unresolved question of whether some marginal bone loss around implants is an immune equilibrium rather than an infection — the foreign-body debate I describe in the osseointegration review — which would change what a millimeter on a radiograph means.

None of that alters what works now. Regular maintenance roughly halves the incidence of peri-implantitis in patients who already have mucositis, screw-retained restorations remove the cement variable entirely, and a baseline record makes the diagnosis a measurement. Those three are within reach of every patient and every practice, and they are where I would put a patient's attention before any new technology.